PTPP: Preference-Aware Trajectory Privacy-Preserving over Location-Based Social Networks

Trajectory privacy-preserving for Location-based Social Networks (LBSNs) has received much attention to protect the sensitive location information of subscribers from leaking. Existing trajectory privacy-preserving schemes in literature are confronted with three problems: (1) it is limited for privacy-preserving by only considering the location anonymization in practical environment, and the sensitive locations are always revealed by this way; (2) they fail to consider the user preference and background information in trajectory anonymization, which is important to keep personalized location-based service; (3) they cannot be adapted to different kinds of privacy risk levels, resulting in low the service precision. To tackle the above problems, we propose PTPP, a preference-aware trajectory privacy-preserving scheme. First, we model the user preference by considering geographical information, semantical information, movement pattern, user familiarity and location popularity. Then, we classify the privacy risk levels according to user familiarity and location popularity. Finally, we propose a preference-aware trajectory anonymization algorithm by considering privacy risk levels. The experimental results show that our method outperforms a state-of-the-art trajectory privacy-preserving method in terms of data utility and efficiency.